Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review
Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review
}
TY - GEN
T1 - Benefits of Mirror Weight Symmetry for 3D Mesh Segmentation in Biomedical Applications
AU - Dordiuk, Vladislav
AU - Dzhigil, Maksim
AU - Ushenin, Konstantin
N1 - This work has been supported by the grant of the Russian Science Foundation, RSF 22-21-00930. The computations were performed on the Uran supercomputer at the IMM UB RAS.
PY - 2023/9/28
Y1 - 2023/9/28
N2 - 3D mesh segmentation is an important task with many biomedical applications. The human body has bilateral symmetry and some variations in organ positions. It allows us to expect a positive effect of rotation and inversion invariant layers in convolutional neural networks that perform biomedical segmentations. In this study, we show the impact of weight symmetry in neural networks that perform 3D mesh segmentation. We analyze the problem of 3D mesh segmentation for pathological vessel structures (aneurysms) and conventional anatomical structures (endocardium and epicardium of ventricles). Local geometrical features are encoded as sampling from the signed distance function, and the neural network performs prediction for each mesh node. We show that weight symmetry gains from 1 to 3% of additional accuracy and allows decreasing the number of trainable parameters up to 8 times without suffering the performance loss if neural networks have at least three convolutional layers. This also works for very small training sets.
AB - 3D mesh segmentation is an important task with many biomedical applications. The human body has bilateral symmetry and some variations in organ positions. It allows us to expect a positive effect of rotation and inversion invariant layers in convolutional neural networks that perform biomedical segmentations. In this study, we show the impact of weight symmetry in neural networks that perform 3D mesh segmentation. We analyze the problem of 3D mesh segmentation for pathological vessel structures (aneurysms) and conventional anatomical structures (endocardium and epicardium of ventricles). Local geometrical features are encoded as sampling from the signed distance function, and the neural network performs prediction for each mesh node. We show that weight symmetry gains from 1 to 3% of additional accuracy and allows decreasing the number of trainable parameters up to 8 times without suffering the performance loss if neural networks have at least three convolutional layers. This also works for very small training sets.
UR - http://www.scopus.com/inward/record.url?partnerID=8YFLogxK&scp=85180377048
U2 - 10.1109/CSGB60362.2023.10329838
DO - 10.1109/CSGB60362.2023.10329838
M3 - Conference contribution
SN - 979-835030797-9
SP - 100
EP - 107
BT - 2023 IEEE Ural-Siberian Conference on Computational Technologies in Cognitive Science, Genomics and Biomedicine, CSGB 2023 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2023 IEEE Ural-Siberian Conference on Computational Technologies in Cognitive Science, Genomics and Biomedicine (CSGB)
Y2 - 28 September 2023 through 30 September 2023
ER -
ID: 50627272